Earthquakes occur as a result of the pressure that builds up between continental plates moving relative to one another in the Earth’s crust. These primary quakes then generate large numbers of aftershocks -- secondary shocks that can occur at some distance from the earthquake's epicentre and also up to a day or two later. Until now, scientists believed that it was changes in the "static stress" brought about by an earthquake that was responsible for its aftershocks since it seemed that only this mechanism could generate aftershocks that take place significantly after the main quake. But Felzer and Brodsky have concluded this is not the case.

The researchers looked at data on the precise locations of thousands of magnitude 2 to 6 earthquakes that occurred in Southern California between 1984 and 2002. They found that the number of aftershocks drops off steeply with increasing distance from the main shock, up to a distance of 50km. More precisely, they found that the frequency of aftershocks decays with distance as a single inverse power law with an exponent of around -1.35. They say this smooth trend means a single triggering process operates over the entire range, and that because static stress is negligible at distances approaching 50 km "dynamic stress" must be the culprit. They also point out that seismic waves also decay quickly with distance following a power-law relation.

The results could also have implications for predicting where aftershocks will occur once an earthquake has started say the researchers: "One of the key predictions from our studies is that the probability of having an aftershock is proportional to the amplitude of the shaking," explains Brodsky. "So if you know the amplitude of the seismic waves coming in you can, in a probabilistic sense, predict whether or not there's going to be an aftershock at a given site."